Document 14076954

advertisement
cs6630 | September 11 2014
COLOR
Miriah Meyer
University of Utah
administrivia . . .
3
-
data exploration assignment due on Tuesday
4
last time . . .
5
MARK TYPES
marks as nodes (items)
points (0D)
lines (1D)
areas (2D)
marks as links
containment
6
connection
expressiveness
(how much)
(what or where)
expressiveness
effectiveness
WHAT’S SO SPECIAL ABOUT THE PLANE?
10
TIME AS ENCODING CHANNEL
!
-
external versus internal memory
- easy to compare views by moving eyes
- hard to compare view to memory of what you saw
11
Get it right in black and white.
Maureen Stone
12
today . . .
13
purpose of color
- to
label (color as a noun)
- to
measure (color as a quantity)
- to
represent and imitate (color as a symbol)
- to
enliven and decorate (color as beauty)
Matterhorn, Landeskarte der Schweiz (Wabern, 1983)
Color
functions of color
identify, group, layer, highlight
e 4.1 Finding the cherries is much easier with color vision.
Colin Ware, Information Visualization: Perception for Design
-
what is color?
-
how do we see color?
-
color deficiencies
-
color spaces
-
guidelines
-
tools
17
what is color?
18
COLOR
the property possessed by an object of
producing different sensations on the eye as
a result of the way it reflects or emits light
!
Oxford Dictionary
19
light
20
(human) visible light
21
Color != Wavelength
but rather, a combination of wavelengths and energy
the role of objects
how do we see color?
24
25
trichromacy
-
possessing three independent channels
for conveying color information
-
-
!
each type of cone contains a specific photosensitive pigment
-
-
derived from three cone types
each pigment is especially sensitive to a certain wavelength of light
!
likelihood of response is both wavelength and intensity based
-
-
thus brain could not distinguish color with input from only one type of
cone
interaction between at least two types of cones is necessary to
perceive color
26
27
opponent-process model
-
-
-
-
trichromatic theory explains how eye receives signals;
opponent process theory explains how signals are processed
!
visual system detects differences between the response of cones
!
three opponent channels
-
red vs green
-
blue vs yellow
-
black vs white (luminance)
+
+
-
!
opposite colors are never perceived together
-
-
no reddish green or bluish yellow
achromatic
28
chromatic
metamers
29
trichromacy
all spectra can be reduced to precisely three values without
loss of information with respect to the visual system
!
!
!
metamerism
any spectra that create the same trichromatic response are
indistinguishable
30
color deficiencies
32
33
color deficiency
-
-
sometimes caused by faulty cones,
sometimes by faulty pathways
!
red-green most common
-
8% of (North American) males, 0.5% of females
!
-
can be explained by opponent color theory
34
35
37
color spaces
38
space of human color
CIE color space
-
-
-
experiments done in the
1920’s and 1930’s
humans can mimic any pure (visible) light by
addition and subtraction of three primary lights
CIE (International Commission on Illumination)
-
standardized a set of color-matching functions that form
the basis for most color measurement instruments
-
-
with RGB, addition and
subtraction were required
to get all visible wavelengths
in nature, light adds (but does not subtract)
any three primaries (additive) can produce only
a subset of all visible colors
R: 645 nm, G: 526 nm, B: 444 nm
CIE color space
Tutorial
Representing
Colors as Three
Numbers
R
GB in graphics is both a way of specifying
color and a way of viewing color. Graphics algorithms manipulate RGB colors, and the images
produced by graphics algorithms are encoded as RGB
pixels and displayed on devices that render these pixels
by emitting RGB light. Colored images are also used to
specify color in graphics. These images may be captured
by cameras or scanners, interactively drawn using tools
such as Adobe PhotoShop, or algorithmically generated. But, what do
How do three numbers, such all of these RGB values mean with
respect to color perception? How
does the RGB triple captured by a
as RGB or XYZ, represent
digital camera relate to the RGB pixels displayed on a monitor? How
color perception, and how
does the RGB triple selected with an
interactive color tool relate to the
are these representations
RGB triple used to color an object in
a 3D rendering?
related to each other and to
Most computer graphics texts and
tutorials provide a description of
physical color? When do
human color vision and measurement as defined by the CIE tristimthey fail?
Editor: Frank Bliss
Maureen C. Stone
StoneSoup Consulting
ulus values, XYZ. Often missing, however, is an in-depth
discussion of the relationship between the different
applications of RGB and XYZ, and any discussion of color
models beyond trichromacy. The goal of this tutorial is
to provide a complete, concise analysis of RGB color
specification and its relationship to perceptual and physical specifications of color, and to introduce some models for color perception beyond tristimulus theory.
RECOMMENDED
Representing color as three numbers
That color can be represented by three numbers—
whether RGB or XYZ—is a direct result of the physiology of human vision. Electromagnetic radiation whose
wavelength is in the visible range (370 to 730 nanometers) is converted by photopigments in the retinal cones
into three signals, which correspond to the response of
the three types of cones. This response is a function of
wavelength and is described by the spectral sensitivity
curves for the cones, as Figure 1 shows.
Colored light can be represented as a spectral distribution, which plots power as a function of wavelength.
(Other fields, such as signal processing, plot spectra as a
function of frequency, which is the inverse of wavelength.) The cones convert this to three cone response
what are the primary colors?
1. red, green, blue
2. red, yellow, blue
3. orange, green, violet
4. cyan, magenta, yellow
what are the primary colors?
1. red, green, blue
2. red, yellow, blue
3. orange, green, violet
4. cyan, magenta, yellow
5. all of the above
paint mixing
-
physical mixing of opaque paints
-
primary: RYB
-
secondary: OGV
-
subtractive
46
ink mixing
-
subtractive mix of transparent inks
-
primary: CMY
-
secondary: RGB
-
-
approx black = C+M+Y
- true black = C+M+Y+K
subtractive
47
light mixing
-
additive mix of colored lights
-
primary: RGB
-
secondary: CMY
-
additive
48
-
very common color space
-
additive, useful for monitors
-
not perceptually uniform
49
D65: midday sun in Western Europe
RGB color space
perceptual color spaces
change in amount of a color value should produce an equivalent visual change
50
HS L|V|B color spaces
-
common cylindrical-coordinate representations
of points in RGB space
-
rearrange geometry of RGB in attempt to be more intuitive
and perceptually relevant
-
hue: what people think of as color
-
saturation: amount of white mixed in
-
luminance: amount of black mixed in
lightness vs value (or brightness)
- intensity, in computer vision applications
-
-
chroma vs saturation
chroma is colorfulness relative to the brightness of another
color that appears white under similar viewing conditions
- saturation is colorfulness of a color relative to its own
brightness
-
51
HS L|V|B color spaces
-
common cylindrical-coordinate representations
of points in RGB space
-
rearrange geometry of RGB in attempt to be more intuitive
and perceptually relevant
-
hue: what people think of as color
-
saturation: amount of white mixed in
-
luminance: amount of black mixed in
lightness vs value (or brightness)
- intensity, in computer vision applications
-
-
chroma vs saturation
chroma is colorfulness relative to the brightness of another
color that appears white under similar viewing conditions
- saturation is colorfulness of a color relative to its own
brightness
-
51
CIE L*a*b* color space
-
designed to approximate human vision
-
describes all colors visible to human eye
-
uses positive and negative values
-
L*: lightness
-
a*: red/magenta to green
-
b*: yellow to blue
-
relative to a point of white (D50)
-
supersedes RGB and CMYK
52
luminance is tricky
53
in this class…
hue
saturation
luminance
guidelines
55
what is a colormap?
-
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
what is a colormap?
-
[0,8]
specifies a mapping between color and values
-
sometimes called a transfer function
categorical vs ordered
- sequential vs diverging
- segmented vs continuous
- univariate vs bivariate
-
-
expressiveness: match colormap to attribute
type characteristics!
56
guidelines
-
-
-
-
ordered colormaps should vary along
saturation or luminance
bivariate colormaps are difficult to interpret if
at least one variable is not binary
categorical colors are easier to remember if
they are nameable
number of hues, and distribution on the
colormap, should be related to which, and
how many structures in the data to emphasize
-
min or max, ends or middle, etc…
57
size & color
58
size & color
“the smaller the mark, the less
distinguishable are the colors”
-Jacques Bertin
58
size & color
59
Small Areas
small areas
© Pfister/Möller
59
which area is bigger,
red or green?
guidelines
-
saturation and hue are not separable in
small regions
-
-
in small regions use bright, highly saturated colors
saturation interacts strongly with size
more difficult to perceive in small regions
- for points and lines use just two saturation levels
-
-
higher saturation makes large areas look
bigger
-
use low saturation pastel colors for large regions
and backgrounds
62
simultaneous contrast
63
simultaneous contrast
63
66
66
luminance contrast
67
guidelines
-
color is a relative medium
-
-
if encoding ordinal data with color, place marks
on solid, neutral background
because of contrast effects, it is difficult to
perceive absolute luminance of
noncontiguous regions
use only 2-4 bins when background is
nonuniform
- for text, ideally use 10:1 ratio, 3:1 minimum
-
68
hues for categories
69
distinguishability
only good at distinguishing 6-12 simultaneous colors
70
Orderthese
Thesecolors…
Colors
order
Based on slide from Stasko
Orderthese
Thesecolors…
Colors
order
Based on slide from Stasko
Orderthese
Thesecolors…
Colors
order
Based on slide from Stasko
guidelines
-
-
luminance and saturation are most
effective for ordinal data because they
have an inherent ordering
hue is great for categorical data because
there is no inherent ordering
-
but limit number of hues to 6-12 for
distinguishability
74
rainbow colormaps: challenges
no implicit order
Borland 2007
rainbow colormaps: challenges
no implicit order
easy to order
Borland 2007
rainbow colormaps: challenges
no implicit order
easy to order
creates artifacts
not perceptually linear
lower resolution
Borland 2007
rainbow colormaps: challenges
zero crossing not explicit
76
rainbow colormaps: challenges
77
rainbow colormaps
RECOMMENDED READING
RECOMMENDED READING
79
guidelines
poor
good
Face-based Luminance Matching for Perceptual Colormap Generation
Gordon Kindlmann∗
School of Computing
University of Utah
Erik Reinhard
School of Electrical Engineering and Computer Science
University of Central Florida
A BSTRACT
Most systems used for creating and displaying colormap-based visualizations are not photometrically calibrated. That is, the relationship between RGB input levels and perceived luminance is usually
not known, due to variations in the monitor, hardware configuration, and the viewing environment. However, the luminance component of perceptually based colormaps should be controlled, due
to the central role that luminance plays in our visual processing.
We address this problem with a simple and effective method for
performing luminance matching on an uncalibrated monitor. The
method is akin to the minimally distinct border technique (a previous method of luminance matching used for measuring luminous
efficiency), but our method relies on the brain’s highly developed
ability to distinguish human faces. We present a user study showing that our method produces equivalent results to the minimally
distinct border technique, but with significantly improved precision.
We demonstrate how results from our luminance matching method
can be directly applied to create new univariate colormaps.
surface shape could come from luminance variations in the univariate colormap itself.
Exerting control of luminance in colormap-based visualizations
is an interesting problem, due to at least three confounding issues. Most importantly, the display device tends to be uncalibrated (proper calibration would require an external measurement
device [7]). The chromaticities, intensities, and response functions
of the primary colors are often not known, and can vary significantly between display devices [13]. Also, the lighting conditions
and configuration of the room are often unknown (or uncontrolled),
contributing to factors such as light reflecting off the display device surface, and differences in brightness and color perception
caused by variations between foveal and peripheral luminous sensitivity [26]. Finally, yellow pigments in the ocular media such as
the lens and the macular area of the retina, can cause non-trivial
differences between individuals in spectral sensitivity [26]. Our experience has been that there is so much variation between monitors
that it is not sensible to simply assume a “standard” monitor and
then work in a CIE colorimetric space such as XYZ, or the approximately perceptually uniform spaces CIELAB and CIELUV.
We address the general problem of colormap luminance control
by proposing a novel technique for luminance matching. Given a
fixed reference color, and a test color with lightness varied by a
user interface, our technique facilitates matching the luminance of
the two colors. The technique is based on the brain’s special capac-
RECOMMENDED READING
CR Categories: I.3.3 [Computing Methodologies]: Computer
Graphics—Picture/Image Generation I.3.4 [Computing Methodologies]: Computer Graphics—Graphics Utilities I.4.10 [Computing Methodologies]: Image Processing and Computer Vision—
Image Representation
Sarah Creem
Dept. of Psychology
University of Utah
Get it right in black and white.
Maureen Stone
82
tools for color
83
84
ColorBrewer palates
85
ColorBrewer palates
sequential
85
ColorBrewer palates
sequential
diverging
85
ColorBrewer palates
sequential
diverging
categorical
85
86
87
88
L7. Intro to Processing
REQUIRED READING
89
Download